Abstract Background Serum CA125 is routinely used in the follow up of ovarian cancer. The objective of the present study was to evaluate the usefulness of CA125 in the detection of ovarian cancer recurrence. Methods This retrospective case study was carried out at a tertiary gynaecological cancer centre in Australia. Patients with all cell types of epithelial ovarian cancer (EOC) treated between 2003 and2010 were considered eligible. We excluded patients whose aim of treatment was palliative, had no follow-up, had no pre-operative CA125 reading or had pre-operative CA125 levels < 35 U/mL. After primary treatment, patients were followed up as per guidelines suggested by National Comprehensive Cancer Network (NCCN). We recorded if symptoms, findings from physical examination, imaging or serum CA125 levels led to the diagnosis of recurrence. An increase in CA125 levels to twice the postoperative nadir was considered as "doubling" at any time during follow up. Results Analysis is based on 56 patients who completed primary treatment and who presented for a total of 274 follow-up episodes. Of those, 29 patients (52%) developed a recurrence within the follow up period. Recurrence was diagnosed by CA125 alone in 14 of 29 patients (48%). CA125 was not elevated in 7 patients (24%) who recurred. Doubling of CA125 from nadir was observed in 27/29 patients. Of those 27 patients the doubling from nadir occurred within the normal range of 35 U/ml in 3 cases and outside the normal range in 24 cases. Multivariate analysis suggests that doubling of serum CA125 (OR 5.10, p 0.036) and nadir CA125 > 10 U/ml (OR 2.86, p 0.01) remained the only independent factors to predict ovarian cancer recurrence. Conclusions The present paper proposes the validation of a novel CA125 algorithm aiming to detect recurrent EOC. These data may allow us to investigate novel ways of follow up that do not require a patient's physical attendance at a clinic (virtual follow-up).